The Empirical Research Between Investor Sentiment And Cross-section Of Stock Return In Chinese Stock Market

When facing to the abnormal phenomenon in the real market, the behavioralfinance works better than the traditional one. And investor sentiment is aways the focusin this field. Until now, the domestic and foreign researchers has achieved the studyabout the aggregate effects of investor sentiment deeply,while knows quite few aboutthe cross-section effects of investor sentiment. So this paper focus on investorsentiment’s cross-section effects to research if investor sentiment is one of determin-ants on the stock’s valuation like Fama-French(1992,1993)and Carhart(1997).On the basis of Baker and Wurgler’s（2006）finding, that the opaque stocks displaygreater sensitivities to investor sentiment, I began the paper. I chose risk, size, age,profitability, growth opportunity, dividends to measure the opacity in the cross-section.In order to make the paper more rigorous, i use two factor model and the extended one,namely, five factor model.Using the Berger and Turtle（2011）model of measurement of stock’s sensitivity toinvestor sentiment, this paper firstly measured all the observed firms’s sensitivities, andthen, classify all the firms into ten portfolios with the sentiment betas increasing acrossportfolios. Next I calculate the firms the time series average for each firm characteristicvariable, and then pool these averages across sentiment sensitivity portfolios to reportthe resultant averages. According to Chinese actual situation, I classified all the firmsinto two categories, one’s sentiment betas is less than zero and the other one’s is biggerthan zero. To the two categories, the results are not the same. In the negative category,the sensitive stocks have high volatility, small size, low PPE (property, plant andequipment), high growth opportunity, low profitability and few dividends, while in thepositive one, it only displays the stocks with high volatility, low age, low PPE and highsale growth are more sensitive to the investor sentiment. And the result is steady in theextended model.After that, I form portfolios to addresses the concern that sentiment proneportfolios may simply be more sensitive. In this step，i designed the equation on thebasis of the CAPM, in order to find if the investor sentiment has the pricingfunction.The result among the negative stocks is satisfactory, and it is consistent withthe former result. However, the positive one show that the relationship between investor sentiment and firm characteristic is faint。What’s more, the result is affirmed by the nextstability analysis, namely, in the negative group, the opaque portfolios, like theportfolios the high volatility, small size, high and low growth opportunity, lowprofitability and few dividends, are affected by investor sentiment obviously, however,the positive group failed to show that kind result.Berk(1995) and Fama and French(2006) provide an excellent discussion of howfirm characteristics maybe related to potential risk factors. This general concern appliesto much of the sentiment literature as the analysis tends to center on unconditionalmeasures of performance without risk corrections, except potentially under anunconditional model. So I learned to Berger and Turtle’s（2011）method to design aconditional α which contain sentiment measures and potential risk to address thisconcern. The result shows that the sentiment prone stocks have marginal performance,which means the sentiment prone portfolios are these high risk portfolios warrantingadditional risk premia. It is in line with CAPM’s Income-Risk Equilibrium Theory.Therefore, I end up this paper with the view that investor sentiment is a determinant ofstocks’s cross-section price.